Hello
excuse me i dont know about pairwise and listwise plz tell me your question clearly.
thanks
Best regards
Elnaz
What is the difference between excluding cases listwise and excluding cases pairwise?
Hello
excuse me i dont know about pairwise and listwise plz tell me your question clearly.
thanks
Best regards
Elnaz
Thanks for replying.
In SPSS, for missing values you can choose to exclude them from any analysis either listwise or pairwise. My question is:
What is the difference between listwise and pairwise??
I'm not sure exactly what the difference is between listwise and pairwise, but I do know the difference between listwise and analysis-by-analysis, which is the choice I often have to make with the analyses I run.
If you exclude cases listwise, then that means that any case that has missing data for any variable included in what you're running will not be included in any of your analyses. So, let's say you're doing a one-way ANOVA with a number of dependent variables. If you exclude cases listwise, SPSS will drop any case with missing data in *one* variable from the analysis of *all* variables, even if that case is only missing data for the one variable.
Excluding cases analysis by analysis means that, if you were running statistics for a number of different dependents, cases would be excluded *only* for the dependents for which they had missing data (rather than being excluded from the whole thing).
I would imagine that excluding cases pairwise is similar -- it lets you keep cases with incomplete data in analyzing any variable(s) where they *do* have data, rather than skipping them entirely (as with excluding listwise). Now, whether or not you want to do that is another matter!
Hello
thanks
i say you , the best work for learning about every thing in SPSS is use of it's help ,
you go in help\ topics\ index
then type your word e.g listwise then click enter ,you can see information about it.
Best regards
Elnaz
Last edited by elnaz; 08-02-2006 at 11:24 PM.
This is an important but subtle difference. When SPSS runs the statistical formulas it calculates various measurements such as means and standard deviations. If there are values missing for a certain variable, and listwise exclusion is used, SPSS will simply not include those variables in these calculations (e.g. of a variable mean). Means, standard deviations, and so forth are then used in regression formulas and ANOVA analysis. For certain situations, however, deleting listwise can misinterpret the data. If the variables are not selected INDEPENDENTLY of each other then you will miss-represent the data by using listwise exclusion.
Say, for example, you're comparing the amount of time that an individual spends on different activities. You have a code specifying the individual and values for time usage variables. Suppose that one of these values is missing. You CANNOT in this situation simply delete that value and calculate the mean for that variable when running a regression (i.e. use listwise exclusion). The variables are not independent; each value corresponds with an individual which corresponds, in turn, with another time-usage value. Because of this, if a value is missing for an individual's time usage variable, all variable values for that individual MUST BE EXCLUDED (i.e. use pairwise exclusion).
BE SURE YOU UNDERSTAND THIS DIFFERENCE AND WHETHER YOUR VARIABLES ARE INDEPENDENTLY SELECTED OR DEPEND ON A DAY/INDIVIDUAL THAT PAIRS THE VALUES. A good rule of thumb is this: if you randomly rearranged one variable's values and kept another variable in original order, would the data still make sense or would things not logically add up? If the answer is no--(e.g. mixing the activity times of multiple individuals will cause an individual to have more than 24 hours of daily activity time)--then you MUST use pairwise exclusion.
Hope that helps.
Last edited by mherman; 12-01-2009 at 08:25 PM.
Tweet |